INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Yo

-0.74

 Horus

-0.73

ossier

-0.73

 Highlands

-0.69

Wonder

-0.69

 Teach

-0.68

 Caribbean

-0.67

pedia

-0.66

Reviewed

-0.65

 Cascade

-0.65

POSITIVE LOGITS

uctor

0.67

 neighb

0.67

tten

0.66

ingen

0.66

 acceler

0.66

itudes

0.65

cules

0.64

 incomp

0.63

icidal

0.63

 separ

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.